When are the machines going to take our jobs?

Posted on 18th September 2019

Adam Thilthorpe

Written by Adam Thilthorpe, Director for Professionalism, BCS, The Chartered Institute for IT

Worry about artificial intelligence taking jobs from human workers is a narrative that’s gaining momentum. Whether it’s call centres being replaced by machines, driverless cars and AI doctors, mainstream headlines are sure of one thing: robots are after our livelihoods. But are these concerns warranted and what can we safely predict about the jobs needed for tomorrow’s society to flourish?

To lift the lid on this topic and other AI related issues, BCS – The Chartered Institute for IT – surveyed its members for their opinions about artificial intelligence.

Right now, there will ‘no job losses’. That’s one of the headline pieces of intelligence BCS gleaned from its member survey. Rather, humans, members believe, can look forward to more augmentation – the idea being that a human worker will operate hand-in-hand with an AI.

That wasn’t a universally held view – but the research’s subtext was clear that we shouldn’t be negative about how AI will change the job market. One member gave an insightful overview: ‘no jobs will be lost – all jobs will be affected and impacted. Some aspects will be automated, not replaced. After all, any rules-based role can be automated.’

Jobs in two years’ time

The BCS research revealed the following jobs might, if they are not already, be affected in the next few years:

  •  Data analysts
  • Testing roles
  • Data entry
  • Helpdesks, service desks, call centre functions, customer first contact, pre-sales expert
  • Recruitment
  • Translation
  • Credit control
  • Compliance violations, security monitoring and threat intelligence

Areas where BCS members thought augmentation would be more quickly seen included intelligent IT security analysis, next generation firewalls, enhanced business intelligence and market intelligence.

Insurance was also cited as a business that may be affected. One member commented: ‘most low-value underwriting decisions will be made by machines. Repetitive clerical/administrative tasks also’. Another cited insurance underwriting: ‘providing automated decisioning and decision-support based on new evidence sources such as health records.’ Another insurance related term that was considered will be affected by AI was FNOL – first notice of loss.

On a broader societal scale, big impact was also expected in medical diagnostics and other pattern recognition-led functions such as exoplanet detection.

Fitting with the theme of AI being an augmentation and enhancement of roles – rather than a source of direct job loss – were these two comments. Firstly: ‘I work in education and there is almost a threat that AI will replace or automate teachers. But I am really not convinced. I think there is opportunity for more automation to reduce workload.’

Secondly: ‘I see AI as providing decision-support for technical roles, not as a replacement for them. As such AI with limited capabilities can be introduced soon (possibly within two years) and then refined and improved in service.’

Jobs affected – the five year view

Further down the line, augmentation was still the watchword. One BCS member wrote: ‘AI can augment the audit and data analysis task, which is likely to improve quality with increasing volumes. It is difficult to see a role in this sector that can be replaced, but it might enable timelier Q&A about legal texts and the like.’

Here is a selection of other roles that BCS members considered likely to be affected within five years:

  • Greater automation on support: in marketing, market intelligence, analysing customers’ behaviour through the internet; and in finance, analysis functions.
  • Increasing automated vehicle control, automated traffic flow management, customs processing and smuggling detection, network traffic flow optimisation. (In April 2019 Elon Musk predicted level 5 self-driving cars by 2020, critics maintain it will be another decade).
  • Automotive, elements of passenger transport, elements of highway traffic control, management / expert decision support tools, data querying and visualisation, linear optimisation problems, call centres, warehouse and dockside goods handling, some logistics, many fail-safe systems.
  • Teachers could possibly use automation more in terms of assessment – and then patterns might be presented in terms of individual pupil progress. There might be improvements in development of programming languages (PL) or pedagogies used in the teaching of computer science which will be because of AI/ML – for example using eye tracking and then ML to compare PL or pedagogical approaches… this is starting to be seen. There could be data crunching and data analysis type opportunities e.g. comparing GCSE results with other local data… this might be used to inform policy.

BCS is a charity with a royal charter, our agenda is to lead the IT industry through its ethical challenges, to support the people who work in the industry, and to make IT good for society.

To read more about BCS research portfolio, visit this link. 

Adam Thilthorpe is speaking at the 2019 Power & Responsibility Summit.

whois: Andy White Freelance WordPress Developer London